AGENTIC-WORKFLOWS
30 days · UTC
Synchronizing with global intelligence nodes...
Production-hardening AI agents: trace, fallbacks, and tenant-safe memory
Teams are shifting from shiny agent diagrams to gritty production work: tracing, offline fallbacks, and memory isolation. [The New Stack](https://the...
Agent routing in LLM systems can cost more than a single-model run
Routing inside LLM agents can raise cost and latency versus running one capable model end-to-end. Avi Chawla argues that naive per-step routing insid...
awesome-ai-agents-2026 ships v0.1.0-alpha: a searchable AI Agent Registry
The awesome-ai-agents-2026 project published v0.1.0-alpha, turning its curated list into a searchable AI Agent Registry. The [v0.1.0-alpha](https://g...
OpenAI’s agentic shift: 3.1 agent-workdays per human day
OpenAI says its research org now runs coding agents at scale, hitting 3.1 agent-workdays per human day. OpenAI framed this as an “automated research ...
Gemini 3.8 Flash vs. Muse Spark 1.3: Long-horizon coding shifts from price tags to token burn
Google Gemini 3.8 Flash and Meta Muse Spark 1.3 both changed long-horizon coding economics by boosting persistence while keeping per-token pricing fla...
Tencent’s Hy4: 1M‑token open‑weight model for repo‑scale agents
Tencent released the open-weight Hy4 LLM with a 1M-token context window, signaling a push toward self-hosted, long-context coding and agent workflows....
Agents need ops: stop running them like scripts
Agent builders are shifting from laptop demos to production-grade services with serverless endpoints, hardened workers, and real spend/health guardrai...
MCP grows up: Docker’s Toolkit and ecosystem updates make agent workflows deployable
Agent infrastructure is consolidating around MCP with real tooling that ops teams can run, monitor, and standardize. Docker put a GUI on MCP inside D...
Multi-turn attacks are beating AI agents — start testing like an adversary
Multi-turn social engineering is quietly defeating AI agents, and a new test method shows why our single-prompt checks miss it. EPFL introduced STING...
Grok 4.6 makes long‑running agents cheaper and sturdier
SpaceXAI launched Grok 4.6 with aggressive pricing and persistent memory aimed at long‑running agents. Early reports say Grok 4.6 cut token prices an...
Google’s Managed Agents aim to cut agent token burn and add hard budget guardrails
Google’s Gemini 3.6 Flash and updated Managed Agents API change how long-running agents spend tokens and handle orchestration. According to this rund...
Perplexity’s take on agent sandboxes: isolating state to make AI work predictable
Perplexity laid out how it isolates AI agents in sandboxed "Spaces" to make stateful workflows safer and more reliable. In this [The New Stack interv...
Agent security shifts from alerts to runtime enforcement
Agent security is moving from alerts to autonomous enforcement, with OpenAI’s Codex Security CLI and new runtime blockers reacting to real prompt-worm...
Safari MCP brings agents into the browser; routing and cost tooling catch up
Safari’s new MCP server lets agents see and act in a real browser, shifting agent work from code-only to runtime debugging. WebKit’s Safari Technolog...
Copilot becomes a native JetBrains agent; Java SDK adds annotation-based tools
GitHub Copilot now runs natively inside JetBrains AI Assistant, and the Java SDK adds annotations for first-class tool wiring. GitHub says Copilot is...
Agents Need a Governance Layer Before They Scale
Agentic AI is stalling on governance, not models or UI, and that changes where backend teams need to invest. An industry brief argues the real bottle...
Agentic-QE ships runtime “oracle” evals, durable-first tests, and a stability layer
Agentic-QE now grades generated tests by running them against real and deliberately-broken code, and locks down its CLI/API behavior. The new release...
Loop Engineering, Not Prompts: How to Make Coding Agents Ship Safely
AI coding agents are moving from prompt hacks to loop engineering with verifiable checks, tighter scopes, and single‑agent workflows that actually shi...
From stack traces to runbooks: making AI agents deterministic and debuggable
AI coding agents are shifting from prompt fiddling to deterministic, documented workflows you can actually debug and operate. [The New Stack](https:/...
From brittle tests to agentic QA: making AI-written code safe to ship
Teams are shifting from manual review and brittle test scripts to agentic QA and pattern-tracking AI code review to keep AI-accelerated releases safe....
OpenAI’s Responses API quietly rewrites how you build long-running agents
OpenAI’s Responses API moves agent workflows to server-side sessions and webhooks, cutting glue code and changing how you run long tasks. A developer...
Agents are now a real workload: Uber’s bill and 1,000 deploys/month force ops, cost, and control rewrites
AI agents moved from experiments to production-scale work, exposing gaps in pipelines, cost controls, and access governance. Teams are hitting deploy...
CodeSignal rolls out agentic AI assessments so you can measure real AI skills
CodeSignal introduced agentic AI assessments that simulate real work so you can measure how your team actually uses AI. The new [Agentic AI Assessmen...